Threads and Interpreter death

[Carl J. Van Arsdall]

I've been toying with threads a lot lately and I've noticed that if a
scripting error occurs in a thread the thread dies, but not the process
that spawned the thread.

Is python supposed to behave this way or is this type of behavior
accidental?

Thanks,

Carl


--

Carl J. Van Arsdall
(e-mail address removed)
Build and Release
MontaVista Software

 

[Rene Pijlman]

Carl J. Van Arsdall:

I've been toying with threads a lot lately and I've noticed that if a
scripting error occurs in a thread the thread dies, but not the process
that spawned the thread.

Is python supposed to behave this way or is this type of behavior
accidental?

"start_new_thread(function, args[, kwargs])
When the function terminates with an unhandled exception, a stack trace is
printed and then the thread exits (but other threads continue to run)."
http://docs.python.org/lib/module-thread.html

 

[Bryan Olson]

Carl J. Van Arsdall:

I've been toying with threads a lot lately and I've noticed that if a
scripting error occurs in a thread the thread dies, but not the process
that spawned the thread.

Is python supposed to behave this way or is this type of behavior
accidental?

"start_new_thread(function, args[, kwargs])
When the function terminates with an unhandled exception, a stack trace is
printed and then the thread exits (but other threads continue to run)."
http://docs.python.org/lib/module-thread.html

And you might also want to see threading.thread's setDaemon() method:

The entire Python program exits when no active non-daemon
threads are left.

http://docs.python.org/lib/thread-objects.html